214 MR. W. F. G. SWANN ON THE SPECIFIC HEATS OF AIR AND 



question, and Qi is the largest flow in the set of experiments. In the case of these 

 experiments on air which were performed at the temperature of the room, the value 

 of A/JSQ! was about yj,, and the value of /i/JSQ 2 about yj,-, so that according to this 

 correction the specific heat, as calculated from the simple equations on pp. 200, 201, 

 would have to be increased by 1'5 parts per 1000.* For the experiments performed 

 at steam temperature the correction is greater in view of the larger value of //, and 

 amounts to an increase of about 3 '2 parts per 1000 on the specific heat.f We meet 

 with another effect at steam temperature, however, which tends to reduce the effect 

 of the above correction. It was observed that the balance point on the platinum 

 thermometer bridge varied with the rate of flow of gas through the apparatus, even 

 when no electric current was passing through the heating coil, the balance point being 

 on the average 17 mm. higher for the small than for the large flow, corresponding 

 to an excess rise in temperature between the thermometers of 0- 034 C. in the 

 small flow. 



The explanation of this phenomenon is not difficult to understand. The gas on 

 entering the calorimeter proper becomes slightly cooled by the leads of the ther- 

 mometer K, so that it arrives at K at a temperature slightly below that of the jacket. 

 As it passes through the central portion of the calorimeter, between the ther- 

 mometers. K and H, it becomes heated up again by the jacket, thus there is a slight 

 dip in temperature, as it were, in the neighbourhood of the thermometer K ; this dip is 

 greater in the small flow than in the large flow, owing to the greater effect of a given 

 loss of heat on the temperature of the smaller quantity of gas. Thus there is a 

 greater rise in temperature between the thermometers in the small than in the large 

 flow. The effect of this dip in temperature is to cause the mean temperature of the 

 gas between the heating coil and the thermometer H to be below that of the jacket 

 by an amount which is greater in the small than in the large flow. The distance 

 from the centre of the thermometer K to the centre of the heating coil was 5 cm., 

 and the distance from the centre of the heating coil to the centre of the thermometer 

 H was 8 cm., so that the alteration in the rise in temperature between the heating 

 coil and this thermometer in the two flows was 0'034x8/13 C. approximately. 

 Thus, if we assume that by the time the gas had reached H it had reattained the 

 temperature of the jacket, we see that the excess of the temperature of the jacket 

 over the mean temperature of the gas between the heating coil and H was approxi- 

 mately 4(0-034) (8/13), or 0'01 C. greater in the small than in the large flow. This 

 fact would tend to minimise the heat loss which took place when the gas became 

 heated by the electric current, the effect being greater in the small than in the large 



* This correction is given more accurately on p. 229. 



t The corrections thus affect the specific heat to a larger extent than that to which they affect the 

 linear relationship between CE/S0 and the flow. The explanation of this is, that whether we take account 

 of the corrections under consideration or not, a straight line fairly accurately represents the relation 

 between CE/S0 and the flow. The slope of the straight line is, however, different in the two cases. 



